Life of a hard drive

[Jetster]

Just ran accost a drive with 52,000 hours and still working
What the most you have seen from a platter drive?

 

[silentbogo]

I've seen a couple of 1TB WD Blue drives with over 60k hours.
Judging from the SMART values, those were used in servers, cause the power cycle count was very-very low.

I think yours is very similar - it was mostly on for several years with minimal stress on the hardware, hence the perfect health even after 6 years .
Maybe it was a backup drive.

 

[jboydgolfer]

I used to see really high values for POH, back when finding 80gb's was far more common than 1tb's. iirc ive had a few 250gb and lower /maxtor drives with 35-45,000+ POH

 

[rtwjunkie]

I've seen a couple of 1TB WD Blue drives with over 60k hours.
Judging from the SMART values, those were used in servers, cause the power cycle count was very-very low.

I think yours is very similar - it was mostly on for several years with minimal stress on the hardware, hence the perfect health even after 6 years .
Maybe it was a backup drive.

Yeah, I've got some server drives approaching 30,000 hours, but with minimal restarts, they are in nearly perfect health as well.

 

[FordGT90Concept]

The majority for me have lived longer than they're useful for (low density/slow speed, not enough capacity). I removed some from my server that were in service 2007-2015 which is 70k hours (replaced by SSDs). They're still functional (just used it a week ago or so for testing a RAID1 in another computer).

I have 3 x 320 GBs that are probably close to 70k, if not passed 70k. All of these drives were in a RAID so I can't check them without pulling them out (not doing that).

Drives that are kept at about 30-40C and don't die in infant mortality period tend to last and last and last.

 

[qubit]

Drives that are kept at about 30-40C and don't die in infant mortality period tend to last and last and last.

...and that aren't knocked about. I constantly see ignorant users at our workplace put laptops on the table with a bump while switched on, which makes me wince. Educating them doesn't help much either. Just because they don't see an immediate failure, they don't accept that they're doing damage.

 

[Vya Domus]

...and that aren't knocked about.

Meh , not sure about that either , I have an old laptop that I have abused the hell out of it and the HDD still works just fine. Just like any other piece of hardware the failure rates follow the usual curve : most failures happen very early or after some fixed amount of years , in between is just random.

 

[FordGT90Concept]

Error rates soar at the slightest provocation. He's not wrong. Just as occupants get tossed about like rag dolls inside of a car hit by a truck, the same thing happens with the head inside of a hard drive. The platters are somewhat resistant to motion because of centripetal force. Hard drives can sense acceleration (especially falls) so they can park the head (get it off the platters) before the deceleration occurs. Nudging a drive comes with little warning because the acceleration and deacceleration occurs quickly.

My drive with 60k+ time on them have almost never moved in operation. If they did, it was as gently as possible.

Laptop cases usually have some tech to absorb small vibrations. Laptops installed in vehicles are often ruggedized which means the hard drive is in a rubber-suspended cradle. When I changed the hard drive in my Systemax SR-15 laptop to an SSD, there were 13 screws that I had to remove (4 screws HDD to carriage, 4 screws attach carriage to rubber bushings, 5 screws hold the cover on).

 

[Jetster]

A hard drive unlike other hardware parts is spinning and moving it around while it is spinning is the quickest way to cause bad sectors. It may recover and move on but its still damaged. Its jsut a matter of time until those bad sectors cause bigger problems

 

[Vya Domus]

A hard drive unlike other hardware parts is spinning and moving it around while it is spinning is the quickest way to cause bad sectors. It may recover and move on but its still damaged. Its jsut a matter of time until those bad sectors cause bigger problems

Error rates soar at the slightest provocation. He's not wrong. Just as occupants get tossed about like rag dolls inside of a car hit by a truck, the same thing happens with the head inside of a hard drive. The platters are somewhat resistant to motion because of centripetal force. Hard drives can sense acceleration (especially falls) so they can park the head (get it off the platters) before the deceleration occurs. Nudging a drive comes with little warning because the acceleration and deacceleration occurs quickly.

My drive with 60k+ time on them have almost never moved in operation. If they did, it was as gently as possible.

You are right , but I still don't think it's a very common cause of failure. In my experience I have had far more HDDs that failed and were inside desktops where obviously they never moved than HDDs inside laptops. That doesn't say much in itself, but it's clear that it's not an event biased toward a certain type of cause.

 

[FordGT90Concept]

Infant mortality is (defective unit/design). Example: BackBlaze found a Seagate model that simply couldn't operate in their server environment having significantly higher failure rates than all the rest.

 

[xkm1948]

A 850MB hard drive in my father's old 486. Last time I went back home it was still functional with Windows95 loaded on it. I even played a round of Red Alert on it while I was there. So yeah, 1993 to 2016 and still runs great.

 

[R-T-B]

A 850MB hard drive in my father's old 486. Last time I went back home it was still functional with Windows95 loaded on it. I even played a round of Red Alert on it while I was there. So yeah, 1993 to 2016 and still runs great.

I had a client send me a 4GB IBM travelstar to recover data from (working perfectly, he just lacked IDE cables). Not quite as impressive, but close.

 

[cornemuse]

Anyone remember the olden days when you had to 'park' the heads before turning off the computer??

 

[eidairaman1]

...and that aren't knocked about. I constantly see ignorant users at our workplace put laptops on the table with a bump while switched on, which makes me wince. Educating them doesn't help much either. Just because they don't see an immediate failure, they don't accept that they're doing damage.

Educate them that continued bumping will result in data loss and money lost

 

[TheRagnarok]

These are from my first media server. Still kicking in a Poweredge T20.

 

[FordGT90Concept]

68000 hours is 7.75 years which would be late 2009. That drive started selling in November 2010 to February 2011. The counter is wrong on that drive or Crystal Disk Info.

 

[Toothless]

 

[FreedomEclipse]

This thread kinda reminds me of that server which had been running for something like 30 years and never powered off even once..

 

[Liviu Cojocaru]

I have a WD Black 640GB back home in Romania since 2007 and it still works I think it has more than 60000h on it

 

[taz420nj]

These are from my server (Plywood).. It has gone through many changes over the years, being upgraded with old parts I came across, but the only thing in it that was brand new were the 4x Hitachi Deskstar 2TB drives (RAID5) - bought them almost 6 years ago. They've been chugging away 24x7 ever since (46,000 hours, only 161 power cycles).. About 3 months ago I filled the array to the brim, and the boot drives (scrounged 2.5" WD Scorpios in RAID1) started making noises, so I shut it down while I planned out an upgrade.. Well I just got me 8 2TB Ultrastars and a PERC H700, so basically I'm going to transfer all my media over to the new RAID6 array then use the old drives for boot (RAID10).

Im actually kind of surprised I got 40,000 hours out of the Scorpios before they started giving me problems. Those I shucked out of old Passport portables - and they don't exactly get the best treatment.

 

[Beastie]

WD blue still going strong. This drive used to be my OS drive and has subsequently been thrashed with torrents, no issues so far.

 

[IBRAHIM_007]

^ old hard drive ( was in my old laptop , overheating alot like reaching 66 c , dropped the laptop some times accidentally in the past )

its nice that its still alive (though caution status ) , i don't trust it at all , i don't use it anymore , only for files that i don't care if it get lost

( almost 51.5 k hours , 44k counts )
51468 hours
44246 count

edit :
sorry i did make a mistake
i mean the opposite
44246 hours = 44k hours
51468 counts = 51k counts